1 /*
2 * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
3 * Copyright (c) 1993
4 * The Regents of the University of California. All rights reserved.
5 *
6 * This code is derived from software contributed to Berkeley by
7 * Paul Borman at Krystal Technologies.
8 *
9 * Copyright (c) 2011 The FreeBSD Foundation
10 * All rights reserved.
11 * Portions of this software were developed by David Chisnall
12 * under sponsorship from the FreeBSD Foundation.
13 *
14 * Redistribution and use in source and binary forms, with or without
15 * modification, are permitted provided that the following conditions
16 * are met:
17 * 1. Redistributions of source code must retain the above copyright
18 * notice, this list of conditions and the following disclaimer.
19 * 2. Redistributions in binary form must reproduce the above copyright
20 * notice, this list of conditions and the following disclaimer in the
21 * documentation and/or other materials provided with the distribution.
22 * 4. Neither the name of the University nor the names of its contributors
23 * may be used to endorse or promote products derived from this software
24 * without specific prior written permission.
25 *
26 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36 * SUCH DAMAGE.
37 */
38
39 #include "lint.h"
40 #include "file64.h"
41 #include <errno.h>
42 #include <limits.h>
43 #include <string.h>
44 #include <stdio.h>
45 #include <stdlib.h>
46 #include <unistd.h>
47 #include <wchar.h>
48 #include "runetype.h"
49 #include "ldpart.h"
50 #include "mblocal.h"
51 #include "setlocale.h"
52 #include "_ctype.h"
53 #include "../i18n/_locale.h"
54
55 /*
56 * A cached version of the runes for this thread. Used by ctype.h
57 */
58 __thread const _RuneLocale *_ThreadRuneLocale;
59
60 extern _RuneLocale *_Read_RuneMagi(FILE *);
61 extern unsigned char __ctype_C[];
62
63 static int __setrunelocale(struct xlocale_ctype *, const char *);
64
65 static int
66 __setrunelocale(struct xlocale_ctype *l, const char *encoding)
67 {
68 FILE *fp;
69 char name[PATH_MAX];
70 _RuneLocale *rl;
71 int saverr, ret;
72 struct xlocale_ctype saved = *l; /* XXX DOUBLE NOT USED */
73 size_t (*old__mbrtowc)(wchar_t *_RESTRICT_KYWD,
74 const char *_RESTRICT_KYWD, size_t, mbstate_t *_RESTRICT_KYWD);
75 size_t (*old__wcrtomb)(char *_RESTRICT_KYWD, wchar_t,
76 mbstate_t *_RESTRICT_KYWD);
77 int (*old__mbsinit)(const mbstate_t *);
78 size_t (*old__mbsnrtowcs)(wchar_t *_RESTRICT_KYWD,
79 const char **_RESTRICT_KYWD, size_t, size_t,
80 mbstate_t *_RESTRICT_KYWD);
81 size_t (*old__wcsnrtombs)(char *_RESTRICT_KYWD,
82 const wchar_t **_RESTRICT_KYWD, size_t, size_t,
83 mbstate_t *_RESTRICT_KYWD);
84 static char ctype_encoding[ENCODING_LEN + 1];
85 static _RuneLocale *CachedRuneLocale;
86 static size_t (*Cached__mbrtowc)(wchar_t *_RESTRICT_KYWD,
87 const char *_RESTRICT_KYWD, size_t, mbstate_t *_RESTRICT_KYWD);
88 static size_t (*Cached__wcrtomb)(char *_RESTRICT_KYWD, wchar_t,
89 mbstate_t *_RESTRICT_KYWD);
90 static int (*Cached__mbsinit)(const mbstate_t *);
91 static size_t (*Cached__mbsnrtowcs)(wchar_t *_RESTRICT_KYWD,
92 const char **_RESTRICT_KYWD, size_t, size_t,
93 mbstate_t *_RESTRICT_KYWD);
94 static size_t (*Cached__wcsnrtombs)(char *_RESTRICT_KYWD,
95 const wchar_t **_RESTRICT_KYWD, size_t, size_t,
96 mbstate_t *_RESTRICT_KYWD);
97
98 /*
99 * The "C" and "POSIX" locale are always here.
100 */
101 if (strcmp(encoding, "C") == 0 || strcmp(encoding, "POSIX") == 0) {
102 int i;
103
104 (void) memcpy(__ctype, __ctype_C, SZ_TOTAL);
105
106 for (i = 0; i < _CACHED_RUNES; i++) {
107 __ctype_mask[i] = _DefaultRuneLocale.__runetype[i];
108 __trans_upper[i] = _DefaultRuneLocale.__mapupper[i];
109 __trans_lower[i] = _DefaultRuneLocale.__maplower[i];
110 }
111
112 (void) _none_init(l, &_DefaultRuneLocale);
113 return (0);
114 }
115
116 /*
117 * If the locale name is the same as our cache, use the cache.
118 */
119 if (CachedRuneLocale != NULL &&
120 strcmp(encoding, ctype_encoding) == 0) {
121 l->runes = CachedRuneLocale;
122 l->__mbrtowc = Cached__mbrtowc;
123 l->__mbsinit = Cached__mbsinit;
124 l->__mbsnrtowcs = Cached__mbsnrtowcs;
125 l->__wcrtomb = Cached__wcrtomb;
126 l->__wcsnrtombs = Cached__wcsnrtombs;
127 return (0);
128 }
129
130 /*
131 * Slurp the locale file into the cache.
132 */
133
134 (void) snprintf(name, sizeof (name), "%s/%s/LC_CTYPE/LCL_DATA",
135 _PathLocale, encoding);
136
137 if ((fp = fopen(name, "r")) == NULL)
138 return (errno == 0 ? ENOENT : errno);
139
140 if ((rl = _Read_RuneMagi(fp)) == NULL) {
141 saverr = (errno == 0 ? EINVAL : errno);
142 (void) fclose(fp);
143 return (saverr);
144 }
145 (void) fclose(fp);
146
147 old__mbrtowc = __mbrtowc;
148 old__mbsinit = __mbsinit;
149 old__mbsnrtowcs = __mbsnrtowcs;
150 old__wcrtomb = __wcrtomb;
151 old__wcsnrtombs = __wcsnrtombs;
152
153 l->__mbrtowc = NULL;
154 l->__mbsinit = NULL;
155 l->__mbsnrtowcs = __mbsnrtowcs_std;
156 l->__wcrtomb = NULL;
157 l->__wcsnrtombs = __wcsnrtombs_std;
158
159 if (strcmp(rl->__encoding, "NONE") == 0)
160 ret = _none_init(l, rl);
161 else if (strcmp(rl->__encoding, "UTF-8") == 0)
162 ret = _UTF8_init(rl);
163 else if (strcmp(rl->__encoding, "EUC-CN") == 0)
164 ret = _EUC_CN_init(rl);
165 else if (strcmp(rl->__encoding, "EUC-JP") == 0)
166 ret = _EUC_JP_init(rl);
167 else if (strcmp(rl->__encoding, "EUC-KR") == 0)
168 ret = _EUC_KR_init(rl);
169 else if (strcmp(rl->__encoding, "EUC-TW") == 0)
170 ret = _EUC_TW_init(rl);
171 else if (strcmp(rl->__encoding, "GB18030") == 0)
172 ret = _GB18030_init(rl);
173 else if (strcmp(rl->__encoding, "GB2312") == 0)
174 ret = _GB2312_init(rl);
175 else if (strcmp(rl->__encoding, "GBK") == 0)
176 ret = _GBK_init(rl);
177 else if (strcmp(rl->__encoding, "BIG5") == 0)
178 ret = _BIG5_init(rl);
179 else if (strcmp(rl->__encoding, "MSKanji") == 0)
180 ret = _MSKanji_init(rl);
181 else
182 ret = EINVAL;
183
184 if (ret == 0) {
185 if (CachedRuneLocale != NULL) {
186 free(CachedRuneLocale);
187 }
188 CachedRuneLocale = l->runes;
189 Cached__mbrtowc = l->__mbrtowc;
190 Cached__mbsinit = l->__mbsinit;
191 Cached__mbsnrtowcs = l->__mbsnrtowcs;
192 Cached__wcrtomb = l->__wcrtomb;
193 Cached__wcsnrtombs = l->__wcsnrtombs;
194 (void) strcpy(ctype_encoding, encoding);
195
196 /*
197 * We need to overwrite the _ctype array. This requires
198 * some finagling. This is because references to it may
199 * have been baked into applications.
200 *
201 * Note that it is interesting that toupper/tolower only
202 * produce defined results when the input is representable
203 * as a byte.
204 */
205
206 /*
207 * The top half is the type mask array. Because we
208 * want to support both legacy Solaris code (which have
209 * mask valeus baked in to them), and we want to be able
210 * to import locale files from other sources (FreeBSD)
211 * which probably uses different masks, we have to perform
212 * a conversion here. Ugh. Note that the _CTYPE definitions
213 * we use from FreeBSD are richer than the Solaris legacy.
214 *
215 * We have to cope with these limitations though, because the
216 * inadequate Solaris definitions were baked into binaries.
217 */
218 for (int i = 0; i < _CACHED_RUNES; i++) {
219 /* ctype can only encode the lower 8 bits. */
220 __ctype[i+1] = rl->__runetype[i] & 0xff;
221 __ctype_mask[i] = rl->__runetype[i];
222 }
223
224 /* The bottom half is the toupper/lower array */
225 for (int i = 0; i < _CACHED_RUNES; i++) {
226 __ctype[258 + i] = i;
227 if (rl->__mapupper[i] && rl->__mapupper[i] != i)
228 __ctype[258+i] = rl->__mapupper[i];
229 if (rl->__maplower[i] && rl->__maplower[i] != i)
230 __ctype[258+i] = rl->__maplower[i];
231
232 /* Don't forget these annoyances either! */
233 __trans_upper[i] = rl->__mapupper[i];
234 __trans_lower[i] = rl->__maplower[i];
235 }
236
237 /*
238 * Note that we expect the init code will have populated
239 * the CSWIDTH array (__ctype[514-520]) properly.
240 */
241 } else {
242 l->__mbrtowc = old__mbrtowc;
243 l->__mbsinit = old__mbsinit;
244 l->__mbsnrtowcs = old__mbsnrtowcs;
245 l->__wcrtomb = old__wcrtomb;
246 l->__wcsnrtombs = old__wcsnrtombs;
247 free(rl);
248 }
249
250 return (ret);
251 }
252
253 int
254 __wrap_setrunelocale(const char *locale)
255 {
256 int ret = __setrunelocale(&__xlocale_global_ctype, locale);
257
258 if (ret != 0) {
259 errno = ret;
260 return (_LDP_ERROR);
261 }
262 /* XXX */
263 // __mb_cur_max = __xlocale_global_ctype.__mb_cur_max;
264 // __mb_sb_limit = __xlocale_global_ctype.__mb_sb_limit;
265 _CurrentRuneLocale = __xlocale_global_ctype.runes;
266 return (_LDP_LOADED);
267 }
268
269 void
270 __set_thread_rune_locale(locale_t loc)
271 {
272
273 if (loc == NULL) {
274 _ThreadRuneLocale = &_DefaultRuneLocale;
275 } else {
276 _ThreadRuneLocale = XLOCALE_CTYPE(loc)->runes;
277 }
278 }
279
280 void *
281 __ctype_load(const char *locale, locale_t unused)
282 {
283 struct xlocale_ctype *l;
284
285 l = calloc(sizeof(struct xlocale_ctype), 1);
286 /* XXX */
287
288 return (l);
289 }